A marketplace of AI skills that actually run: 593 skills ship inside the Nexus Runner desktop app, reading your files, processing them, and producing charts and conclusions on your own machine. It executes, it doesn't just chat.

How is this different from Zapier or just asking a large model?

Zapier connects cloud apps through APIs; we work on the files on your machine. A model hands you prose; we hand you a table, a chart and a conclusion you can check. A skill is a program with an input contract, not a prompt.

No. Skills execute on your computer, source files never leave it, and output is written straight to your folder. Content is sent to the built-in model only when you use AI chat or a skill that needs a model.

What do I need to install?

Just Nexus Runner — the runtime ships with the app. No Python setup, no code.

How much is the free tier?

200 Nex per month on signup, reset at the start of each month, with every skill available and no tier gates. The four Runner tutorial skills are not billed.

How many runs does Pro at CNY 49 cover?

3,000 Nex per month. Skills bill per run and most cost 3-10 Nex, which works out to several hundred runs. Each skill's exact price is on its detail page.

Which Windows versions are supported?

64-bit Windows 10 and Windows 11. Windows is the only desktop client we ship.

How do I connect it to Claude Desktop or Cursor?

Nexus Runner speaks MCP: register it as an MCP server in your client and the skills show up as tools. The docs page walks through the setup step by step.

Where do skills come from? Can I build my own?

Skills are maintained by the platform and its creators, each with an input contract and sample data. You can also build your own in the Runner workbench — those stay yours, never published or encrypted.

Do you offer refunds?

Subscriptions can be refunded under our refund policy; the conditions and process are on the refund policy page. Nex already consumed is not refundable.

◆

API Quality

API quality: env setup to API test to profiling.

Start free Browse skills

Start free — 200 Nex/month

Skills in this bundle

1/3

Config Forge

Generates common DevOps configs: Dockerfile, GitHub Actions workflow, Kubernetes manifests, Nginx site config and systemd units from a runtime, app, domain or service name, with JSON options for details.

Skill5 Nex

2/3

API Warden

One-command composite: OpenAPI metrics + contract validation + design review -> governance report. No LLM. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Skill8 Nex

3/3

Code Vitals

One-command composite audit: structure + complexity hotspots + smells + tech-debt -> unified health report. No LLM. (Action Co-pilot: generates next-best-actions (actions.json), follow-up tasks and calendar reminders on top of the analysis — what to act on today.)

Skill8 Nex

See it in action

Download demo video (MP4)

Quick answer

API quality: env setup to API test to profiling. MCP Nexus orchestrates 3 skills (Env Concierge, API Probe, Py Profiler) into one reproducible workflow that runs on your machine, keeps data in your network, and delivers results in minutes.

Env Concierge: Parse and analyze .env file content: extract key-value pair…
API Probe: Send HTTP requests directly from chat and inspect the respo…
Py Profiler: Parse Python cProfile output and generate structured perfor…

What is API Quality with MCP Nexus

API Quality chains 3 skills (Env Concierge, API Probe, Py Profiler) into one reproducible workflow covering the key steps from input to result, running fully on-premises with data staying in your network, for teams that need API quality: env setup to API test to profiling. on demand.

How it works

1Step 1

Env Concierge

Parse and analyze .env file content: extract key-value pairs, detect empty or sensitive variables,…

2Step 2

API Probe

Send HTTP requests directly from chat and inspect the response. Supports GET, POST, PUT, DELETE, PA…

3Step 3

Py Profiler

Parse Python cProfile output and generate structured performance bottleneck analysis report with op…

MCP Nexus vs alternatives

	MCP Nexus	Manual / Excel	Generic AI / Cloud tools
Pricing	Free tier, metered by calls	Many manual hours	Per-seat plus per-volume fees
Local execution	Yes, data stays on-premises	Manual process	No, data leaves your network
Setup time	Minutes to first run	Days, experience-dependent	~1 week integration and training

Related use cases

◆Code Quality Audit Workflow

<>Engineering Quality & Review

◆Code Quality Gate

◆Dev Toolchain

Start free Browse skills